Download Distribution, aggregate demand and productivity growth: theory and

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Business cycle wikipedia , lookup

Production for use wikipedia , lookup

Chinese economic reform wikipedia , lookup

Productivity improving technologies wikipedia , lookup

Economic growth wikipedia , lookup

Ragnar Nurkse's balanced growth theory wikipedia , lookup

Transformation in economics wikipedia , lookup

Transcript
Cambridge Journal of Economics 2010, 34, 727–754
doi:10.1093/cje/bep066
Advance Access publication 10 November 2009
Distribution, aggregate demand and
productivity growth: theory and empirical
results for six OECD countries based on
a post-Kaleckian model
Empirical research based on the Bhaduri/Marglin-variant of the Kaleckian model
has recently shown that aggregate demand in many medium-sized and large open
economies tends to be wage-led in the medium to long run, even in a period of
increasing globalisation. In this paper we extend this type of analysis and integrate
the effects on productivity growth, theoretically and empirically. Productivity growth
is introduced into the theoretical model making use of the Verdoorn effect or of
Kaldor’s technical progress function and hence of a positive relationship between
GDP or capital stock growth and productivity growth. Further on, a cost-push or
Marx/Hicks-effect and hence a positive impact of real wage growth or the wage share
on productivity growth is taken into account. In the empirical part we estimate
productivity growth equations for six countries introducing these two effects.
Finally, economic policy conclusions are drawn.
Key words: Demand-led growth, Endogenous technical change, Wage-led and profitled demand regimes, Productivity regime
JEL classifications: E12, E21, E22, E25, O41
1. Introduction
The recent two and a half decades have seen remarkable changes in functional income
distribution, in particular in continental European countries (see Table 1). Whereas the
labour income share had a slight tendency to increase until the recession of the early 1980s
it has shown a remarkable downward trend since then. The USA has seen similar but not so
pronounced developments, whereas in the UK functional income distribution has changed
Manuscript received 13 October 2008; final version received 17 May 2009.
Address for correspondence: Prof. Dr Eckhard Hein, Berlin School of Economics and Law, Badensche Str.
50-51, 10825 Berlin, Germany; email: [email protected]
* Berlin School of Economics and Law, Germany (EH) and Leeds University Business School, UK and
University of Hamburg, Germany (AT). For helpful comments we would like to thank the participants in
a seminar on ‘Empirical work based on Kaleckian models’ at the Macroeconomic Policy Institute (IMK),
Hans Boeckler Foundation, Duesseldorf, 2008, and in the 12th conference of the Research Network
Macroeconomics and Macroeconomic Policies, Berlin, 2008. We are also most grateful for comments and
suggestions by Till van Treeck and Robert Vergeer, and in particular by two anonymous referees. Remaining
errors are, of course, ours.
Ó The Author 2009. Published by Oxford University Press on behalf of the Cambridge Political Economy Society.
All rights reserved.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Eckhard Hein and Artur Tarassow*
728
E. Hein and A. Tarassow
Table 1. GDP growth, productivity growth, real wage growth and labour income share on average over
the business cycle in Austria, France, Germany, the Netherlands, the UK and the USA, 1960–2007 (all
figures are percentages)
Growth of real
labour productivityd
Growth of
real compensation
per employee
Labour
income sharee
4.18
4.69
2.35
2.68
2.32
2.43
4.61
4.29
2.43
2.18
1.89
1.50
5.06
4.88
1.46
2.29
0.62
0.46
80.65
79.77
80.01
74.88
69.62
63.40
5.37
4.29
2.82
2.04
2.23
2.02
4.92
3.49
2.46
2.02
1.26
1.47
5.32
4.39
2.21
1.01
1.19
1.48
73.35
72.50
75.96
70.73
66.64
66.19
3.78
3.74
2.41
2.70
1.56
1.54
3.93
3.54
1.87
1.80
2.11
1.64
4.72
5.36
1.13
1.35
1.44
-0.21
68.25
69.20
70.28
66.83
65.79
63.23
4.47
4.44
1.58
2.72
3.14
1.96
3.06
4.15
1.70
1.53
1.40
1.71
6.03
6.04
0.78
0.53
0.90
0.85
67.29
72.58
74.93
68.78
66.87
65.95
2.87
2.77
1.36
2.27
2.74
2.76
1.97
2.87
1.20
1.90
2.09
1.91
2.40
3.56
1.73
2.06
1.62
2.35
72.87
74.20
75.20
74.31
72.93
72.76
4.22
3.54
2.32
3.47
3.40
2.63
2.30
1.54
0.84
1.44
1.63
1.94
2.67
1.50
0.88
0.76
1.54
1.66
69.89
70.83
69.54
68.41
67.46
66.49
Notes: The local minimum of real GDP growth defines the end year of a business cycle.
Incomplete cycle.
b
Until 1991 only West Germany, from 1991 onwards unified Germany.
c
At 2000 market prices.
d
Real GDP per person employed (equal to full-time equivalents).
e
Compensation per employee as percentage of GDP at factor cost per person employed.
Source: European Commission (2007), authors’ calculations.
a
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Austria
1961–1967a
1968–1975
1976–1984
1985–1993
1994–2002
2003–2007a
France
1961–1968a
1969–1975
1976–1981
1982–1993
1994–2003
2004–2007a
Germanyb
1961–1967a
1968–1975
1976–1982
1983–1993
1994–2003
2004–2007a
The Netherlands
1961–1966a
1967–1975
1976–1982
1983–1993
1994–2002
2003–2007a
UK
1961–1966a
1967–1974
1975–1980
1981–1991
1992–2002
2003–2007a
USA
1961–1970
1971–1974
1975–1982
1983–1991
1992–2001
2002–2007a
Growth of
real GDPc
Distribution, aggregate demand and productivity growth
729
1
See Hein and Vogel (2008) for an overview.
This would require a more detailed analysis of structural change within an economy, of productivity
catching-up processes among economies, as well as productivity growth-conducive institutions. See for
instance the studies by Cornwall and Cornwall (2002) and Vergeer and Kleinknecht (2007).
2
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
even less over the last five decades. These developments beg the question as to their
medium- to long-run effects on aggregate demand, gross domestic product (GDP) growth,
capital accumulation and productivity growth. Starting with the work by Bowles and Boyer
(1995), Bhaduri and Marglin’s (1990) seminal theoretical conclusion that aggregate
demand and capital accumulation may be either wage-led or profit-led in the medium to
long run, also within a Kaleckian demand-driven distribution and growth model, has
increasingly inspired empirical research on the relationship between functional income
distribution and aggregate demand, respectively, growth.1
However, modern Kaleckian distribution and growth models, as well as the empirical
estimations based on the Bhaduri and Marglin (1990) model, have only occasionally taken
productivity growth into account. Rowthorn (1981), Taylor (1991, pp. 225–8), You
(1994), Cassetti (2003) and Dutt (2003) have introduced endogenous productivity growth
into different variants of the Kaleckian model. And Naastepad (2006) has included
productivity issues into her estimations for the Netherlands based on the Bhaduri–Marglin
model, albeit in an incomplete way: There is no effect of productivity growth on investment
in her model, and real wage growth instead of distribution is taken to be the exogenous
variable. This implies that in her model with exogenous real wage rate growth, productivity
growth only feeds back on output growth through its effects on the profit share, but has no
direct effect on investment. Below we will also argue that introducing the wage share (or
the profit share) as cost push factor into the productivity function is more plausible than
using real wage growth.
Our paper builds on the recent theoretical and empirical literature on productivity
growth in Kaleckian models. We attempt to integrate productivity growth into the seminal
Bhaduri–Marglin (1990) model in a consistent and systematic way. In our theoretical and
empirical analysis, the profit share will be considered to be the exogenous variable and
aggregate demand, capital accumulation and productivity growth will be determined
endogenously. Following Setterfield and Cornwall (2002) and also Naastepad (2006), we
will distinguish between the demand regime and the productivity regime in our model. We
will discuss the separate effects of changes in the profit share on each of these regimes, and
finally we will analyse the overall effects of changes in distribution on aggregate demand,
capital accumulation and productivity growth. The model presented below will therefore
be a partial model of a private open economy. Feedback effects of accumulation and
productivity growth on distribution will not be considered. It should also be noted that we
do not claim to deliver a full theory of technical progress but rather limit ourselves to the
effects of distribution on aggregate demand and productivity growth.2 Nonetheless,
extending the recent work in the Kaleckian tradition, we not only touch on the short- to
medium-run demand effects of a falling wage share, but we contribute to an understanding
of the long-run effects on capital accumulation, productivity growth and, hence, on the
potential or the ‘natural rate of growth’. We show that with the endogeneity of productivity
growth, potential GDP becomes path dependent with respect to distributional changes,
and economic policies thus have long-lasting effects through this channel.
Since a lot of empirical work has already been done on the demand regimes in developed
Organisation for Economic Cooperation and Development (OECD) countries, our
empirical estimations will focus on the productivity regime only. We will estimate different
730
E. Hein and A. Tarassow
2. Some stylised facts on GDP growth, productivity growth and distribution
Taking a look at the development of average values over the business cycle since the early
1960s until 2007, some remarkable parallels between GDP growth and labour productivity
growth become obvious, in particular in the continental European countries (Austria,
France, Germany, the Netherlands) (Table 1). The break in the long-run development of
GDP in the mid 1970s, from the high growth rates of the ‘golden age’ to mediocre growth
trend in the ‘post-golden-age’ is accompanied by a similar break in productivity growth
rates, which have shown remarkably lower values since the second cycle of the 1970s than
before. Also real wage growth has dropped significantly in the continental European
countries since the mid 1970s, and since the early 1980s the labour income share has
shown a continuously falling trend.1
The development in the USA and the UK deviate from this almost uniform pattern in
the continental European countries. These two countries also witnessed a drop in GDP
growth and productivity growth during the cycle of the second half of the 1970s. However,
GDP growth and productivity growth had already started to recover in the 1980s and
almost reached the growth rates of the early 1970s again. For real wage growth, which also
dropped in the mid 1970s, in the USA it took until the 1990s to recover, whereas in the UK
the recovery started in the 1980s. The fall of the labour income share since the mid 1970s
(USA) or early 1980s (UK) was therefore less marked than in the continental European
countries.
From these broad developments, the general picture emerges that GDP growth and
productivity growth rates display a close positive relationship. Real wage growth and
productivity growth also seem to be positively related, whereas there seems to exist
a negative relationship between profit share and productivity growth, particularly in the
continental European countries, since the early 1980s.
1
Note that in Table 1 labour productivity growth is for full-time equivalents, and hence reflects hourly
productivity growth, whereas growth of real compensation per employees is not for full-time equivalents.
Since working hours per employee have changed over time, growth of real compensation per employee falling
short of (exceeding) labour productivity growth is not necessarily accompanied by a falling (rising) labour
income share.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
productivity equations, in particular an equation that is derived directly from the
theoretical model. Using the country sample that Hein and Vogel (2008) used in the
estimations of demand regimes, this paper can be seen as an extension of that study. The
USA as a large, rather closed economy, Germany, France and the UK as medium-sized
and increasingly open economies, and Austria and the Netherlands as small open
economies, will be considered.
The remainder of this paper is organised as follows. In Section 2 we briefly review some
stylised facts on GDP growth, productivity growth and distribution for the six countries in
our data set. Section 3 will continue with a simple theoretical model for a private open
economy with technical progress. Section 4 is dedicated to a brief review of recent
empirical studies on the demand regime in the six countries under investigation and
a review of recent estimations of productivity equations in the literature. In Section 5 we
will present the estimation results for our productivity equations, which allow us to draw
some conclusions regarding the productivity regime. Section 6 derives some results
regarding the overall regime in the six countries under investigation and draws some brief
economic policy conclusions.
Distribution, aggregate demand and productivity growth
731
3. The theoretical model
3.1 The demand regime
In order to analyse the effects of changes in distribution on economic activity and capital
accumulation, we start with the goods market equilibrium condition for an open economy
without economic activity of the state in real terms: planned saving (S) has to be equal to
1
See Bhaduri (2006) on a model with endogenous productivity and real wage growth, and hence with
endogenously determined functional income distribution.
2
For more extensive treatment of technical progress in distribution and growth models see, for example,
Foley and Michl (1999) and Kurz (1991).
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The theoretical model is based on the open economy Kaleckian distribution and growth
models by Bhaduri and Marglin (1990) and Blecker (1989). Since in these models there is
no technical change, and hence no productivity growth, we will introduce technical
change into this basic model in three steps, following a procedure suggested by
Setterfield and Cornwall (2002) and applied by Naastepad (2006). In the first step, for the
discussion of the demand regime, i.e. the effects of changes in functional income
distribution on the goods market equilibrium, we assume productivity growth to be
exogenous. In the second step, we will deal with the productivity regime and technical
progress will be endogenised, i.e. we will integrate the effects of changes in income
distribution on productivity growth. In the third step, demand and productivity regimes
will be integrated and we will discuss the effects of changes in income distribution on the
overall regime. In the model, functional income distribution is considered to be the
exogenous variable, determined by institutional factors and relative powers of capital and
labour, i.e. by competition between firms in the goods market and between firms and
labourers in the labour market. The goods market equilibrium rates of capacity utilisation,
profit and capital accumulation, together with the rate of productivity growth, are
endogenously determined. Potential feedbacks from goods market activity and productivity growth to income distribution are excluded from the analysis.1 This procedure
allows our work, with a focus on the productivity regime in the empirical part, to be
connected to the recent empirical studies on the demand regime, as will be seen in Sections
4 and 5.
In order to simplify the following discussion we assume that technical progress is
labour saving and capital embodied. Technical progress is hence associated with a falling
labour–output-ratio (l 5 L/Y) and rising labour productivity (y 5 Y/L). The capital–
labour ratio (k 5 K/L) increases at the same rate as labour productivity does, and the
capital–potential output ratio (v 5 K/Yp), therefore remains constant. This means that we
assume Harrod-neutral technical progress, as in Rowthorn (1981), Cassetti (2003) and
Dutt (2003).2
We deal with an open economy without economic activity of the state, which depends
on imported inputs for production purposes, and the output of which competes in
international markets. Neither the movement of labour nor of capital is considered. We
take the prices of imported inputs and of the competing foreign final output to be
exogenously given and to be moving in step. The nominal exchange rate, the price of
a unit of domestic currency in foreign currency, is determined by monetary policies and
international financial markets, and is also considered to be exogenous for our
purposes.
732
E. Hein and A. Tarassow
net investment (I) plus net exports (NX), the difference between exports (Ex) and imports
(Im) of goods and services:
S 5 I 1 ðEx 2 ImÞ:
ð1Þ
For convenience, equation (1) is normalised by the real capital stock (K). Therefore, we get
the following goods market equilibrium relationship between the saving rate (s 5 S/K),
the accumulation rate (g 5 I/K) and the net export rate (b 5 NX/K):
s 5 g 1 b:
ð2Þ
SW
s 5 SP 1
5 sP P 1 sWKðY 2 PÞ 5 ½sW 1 ðsP 2 sW Þhuv;
K
0 # sW < sP # 1:
ð3Þ
Investment is modelled according to Bhaduri and Marglin (1990): capital accumulation
is a positive function of the profit rate, which can be decomposed into the profit share, the
rate of capacity utilisation and the capital–potential output-ratio (r 5 hu/v). We also
include technical progress, which for the time being is assumed to be exogenous, into the
investment function. Since technical progress is embodied in capital stock it will stimulate
investment. Firms have to invest in new machines and equipment in order to gain from
productivity growth, which is made available by new technologies. This effect on
investment will be the more pronounced the more fundamental technical change is: the
invention of new basic technologies will have a stronger effect on real investment than
marginal changes in technologies already in existence.2
Since the capital–potential output-ratio is assumed to be constant with technical change,
capital accumulation is positively affected by the profit share, indicating unit profits, the
rate of capacity utilisation, indicating (expected) demand, and by productivity growth (ŷ).
In order for domestic capital accumulation to be positive, the expected rate of profit has to
exceed a minimum rate (rmin), given by the foreign rate of profit or by the rate of interest in
financial markets. Both possible minimum rates are considered to be exogenous in the
present model.3
1
According to Kalecki (1954: 28-41) the profit share is determined by the mark-up on unit variable costs
in firms’ pricing in incompletely competitive goods market and by the ratio of (imported) unit material costs
to unit labour costs. The mark-up depends on the degree of competition in the goods market, the relative
importance of price competition, the development of overhead costs and the power labour unions.
2
Dutt (2003) also discusses potential effects of technical progress on saving—new products and hence
consumption possibilities may cause a reduction in the propensity to save—and on the mark-up and hence
income distribution—technology leaders may increase their mark-ups and hence the profit share for the
economy as a whole. We will not integrate these effects into our model.
3
In post-Keynesian models the long-term rate of interest in financial markets is mainly determined by the
central bank’s interest rate policy in the long run. It is therefore an exogenous variable for income generation
and growth whereas the volume of credit and the quantity of money are endogenously determined (Hein,
2008, pp. 30–55, 68–81).
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Saving consists of saving out of profits (SP) and saving out of wages (SW). The propensity
to save out of wages (sW) is assumed to fall short of the propensity to save out of profits (sP),
in particular because the latter includes retained earnings of firms. The profit share relates
profits to domestic income consisting of wages and profits [h 5 P/(W 1 P) 5 P/Y],1 the
rate of capacity utilisation is the relation of output to potential output (u 5 Y/Yp) and the
capital–potential output ratio relates the capital stock to potential output (v 5 K/Yp).
Thus, we obtain for the saving rate:
Distribution, aggregate demand and productivity growth
733
g 5 a 1 bu 1 th 1 vŷ;
ð4Þ
a; b; t; v > 0;
g > 0 for r > rmin :
The net export rate is positively affected by international competitiveness, provided that the
Marshall–Lerner condition can be assumed to hold and the sum of the price elasticities of
exports and imports exceeds unity. Under this condition, the real exchange rate (er) will have
a positive effect on net exports. But net exports also depend on the relative developments of
foreign and domestic demand. If domestic demand grows at a faster rate than foreign
demand, net exports will decline, ceteris paribus. Therefore, an increase in the domestic rate
of capacity utilisation will have a negative impact on net exports, ceteris paribus.
b 5 cer ðhÞ 2 fu;
c; f > 0:
ð5Þ
Stability of the goods market equilibrium requires that saving responds more elastically
towards a change in the endogenous variable, the rate of capacity utilisation, than
investment and net exports do together:
@s @g @b
2 2 >0
@u @u @u
0
1
½sW 1 ðsP 2 sW Þh 2 b 1 f > 0:
v
ð7Þ
We shall only consider stable goods market equilibria and the effects of changes in
distribution on these equilibria. The equilibrium rates (*) of capacity utilisation and capital
accumulation are given by:2
u* 5
a 1 th 1 vŷ 1 cer ðhÞ
;
½sW 1 ðsP 2 sW Þh1v 2 b 1 f
ð8Þ
1
See the results in Bowles and Boyer (1995), Hein et al. (2006), Ederer and Stockhammer (2007),
Naastepad and Storm (2007), Ederer (2008), Stockhammer and Ederer (2008), Hein and Vogel (2008,
2009), Stockhammer, Onaran and Ederer (2009) and Stockhammer, Hein and Grafl (2009).
2
The equilibrium rate of profit can easily be calculated inserting equation (8) into the definition of the
profit rate: r 5 hu/v. However, the profit rate will not be considered explicitly in what follows.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The real exchange rate, which is determined by the nominal exchange rate (e) and by the
relationship between foreign prices ( pf) and domestic prices ( p)—er 5 epf/p—is affected by
changes in the profit share but in an ambiguous way, as has been shown in detail by Hein and
Vogel (2008). Assuming that firms set prices according to a mark-up on unit variable costs,
consisting of imported material costs and labour costs, a change in the profit share can either
be caused by a change in the mark-up or by a change in the ratio of unit costs of imported
materials to unit labour costs (z). If an increase in the profit share is caused by a rising markup, ceteris paribus, domestic prices will rise and the real exchange rate and hence
international competitiveness will decline. But if an increasing profit share is triggered by
a rising ratio of unit imported material costs to unit labour costs, ceteris paribus, the real
exchange rate will increase and international competitiveness will improve. Nominal
depreciation of the domestic currency, i.e. an increase in the nominal exchange rate, or
falling nominal wages will increase the ratio of unit material costs to unit labour costs, and
will therefore make an increasing profit share go along with improved competitiveness.
Empirically, if there is any relationship between the profit share and international
competitiveness, this relationship seems to be positive,1 so that we assume in what follows:
@er
$ 0:
ð6Þ
er 5 er h ;
@h
734
E. Hein and A. Tarassow
½sW 1 ðsP 2 sW Þh1v 1 f a 1 th 1 vŷ 1 bcer ðhÞ
g* 5
:
½sW 1 ðsP 2 sW Þh1v 2 b 1 f
ð9Þ
The effect of a change in the profit share on the rates of capacity utilisation and capital
accumulation can be calculated from equations (8) and (9):
r
t 2 ðsP 2 sW Þuv 1 c@e
@u*
@h
5
;
@h ½sW 1 ðsP 2 sW Þh1v 2 b 1 f
r
@g* t ½sW 1 ðsP 2 sW Þh1v 1 f 2 bðsP 2 sW Þuv 1 bc@e
@h
5
:
1
@h
½sW 1 ðsP 2 sW Þhv 2 b 1 f
ð80 Þ
ð90 Þ
3.2 The productivity regime
Within post-Keynesian distribution and growth theory, Kaldor in particular developed
different ways to endogenise technological change. In his technical progress function
(Kaldor, 1957, 1961), productivity growth is positively affected by the growth of capital
intensity, because technical progress is capital embodied. Another possibility was proposed
by Kaldor (1966) in looking for an explanation of the (slow) growth in the UK. There he
applied Verdoorn’s Law.2 According to Verdoorn (1949), the growth rate of labour
productivity in industrial production is positively associated with the growth rate of output.
This can be explained by static and dynamic economies of scale: the expansion of aggregate
demand, sales and hence the market allows for increasing rationalisation and mechanisation and favourably affects technical progress and productivity growth.
Following these approaches implies that the growth rate of labour productivity is positively
affected by the dynamics of output and/or capital stock. Therefore, we can either integrate
1
Bhaduri and Marglin (1990) also derive an antagonistic–stagnationist regime in which a rising profit
share reduces equilibrium capacity utilisation but increases equilibrium capital accumulation. As can be seen
from equations (8’) and (9’) such a regime arises if investment decisions are hardly affected by capacity
utilisation and hence b is very small.
2
On Verdoorns’s Law see the contributions in McCombie et al. (2002A).
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Assuming the goods market equilibrium to be stable, equation (8’) shows that an
increasing profit share will have no unique effect on equilibrium capacity utilisation: there
are positive effects via investment (t) and net exports [cð@er =@hÞ] but also a negative effect
via consumption [2ðsP 2sW Þðu=vÞ]. For equilibrium capital accumulation a similar result is
obtained, as can be seen in equation (9’). Again, we have positive effects via investment
[tf½sW 1ðsP 2sW Þhð1=vÞ1fg] and net exports [bcð@er =@hÞ] but a negative effect via
consumption [2bðsP 2sW Þðu=vÞ]. Depending on the relative strength of each of these
effects, a rising profit share may cause rising rates of capacity utilisation and capital
accumulation or falling rates. In the first case we would obtain a profit-led demand regime,
in the second case demand would be wage-led. A wage-led regime becomes more likely the
lower the elasticity of investment with respect to the profit share, the lower the effect of
redistribution is for international competitiveness and net exports and the higher the
difference in saving propensities out of profits and out of wages.
Contrary to Bhaduri and Marglin (1990) we abstain from further differentiating the
potential demand regimes with the help of equations (8’) and (9’), but rather continue with
the discussion of the productivity regime.1
Distribution, aggregate demand and productivity growth
735
or:
ŷ 5 h 1 eg 2 uh;
h; e; u > 0:
ð10BÞ
In order to facilitate graphical analysis in the following section, equation (10A) contains
a kind of Verdoorn relationship between output/capacity utilisation and productivity growth,
whereas equation (10b) is reminiscent of the technical progress function, i.e. it includes
a positive relationship between capital stock growth and productivity growth.3 Productivity
growth is hence positively affected by capacity utilisation or capital stock growth, and
negatively by the profit share. Equation (10B) is also used by Cassetti (2003), whereas
Naastepad (2006), in her empirical model for the Netherlands, has chosen real wage
growth, and not the profit or the wage share, as determinant of wage-induced productivity
growth. We follow Cassetti (2003), because we hold that real wage growth will only give an
additional push to capitalists’ efforts to implement technical progress if it exceeds
productivity growth and downward pressure on the profit share or on unit profits is exerted.
Different from the demand regime, a change in the profit share has a uniquely inverse
effect on the productivity regime:
@ ŷ
5 2 u < 0:
ð10a; b0 Þ
@h
Independently of capacity utilisation, capital stock growth or income distribution,
productivity growth in equation (10A) or equation (10B) is also affected by a constant
1
See also Lima (2004) who makes use of a non-linear effect of the wage share on technological innovations
in a somewhat more complex model than ours. However, in his model there is no Verdoorn effect or technical
progress function (see also Lima, 2000).
2
In a Kaleckian model of an open economy, as the one presented here, nominal wage growth exceeding
productivity growth will cause a rise in the wage share and a drop in the profit share, even if the mark-up on
unit labour costs in firms’ pricing remains constant (Hein, 2005).
3
Since the relationship between the goods market equilibrium rates of capacity utilisation and capital
accumulation is given by equations (8) and (9), this also implies a definite relationship between the
coefficients r and e in equations (10A) and (10B). This is of no further importance for the following
qualitative theoretical analysis.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
capacity utilisation or capital accumulation into the equation determining productivity
growth. Rowthorn (1981) and Dutt (2003), for example, have chosen the latter way of
integration productivity growth into Kaleckian distribution and growth models.
Apart from aggregate demand and output, we will consider a second determinant of
productivity growth, which has been taken into account in recent theoretical and empirical
work on the basis of the Kaleckian model. Cassetti (2003) and Naastepad (2006), as well as
Taylor (1991, pp. 225–8), have introduced a wage push variable into the productivity
equations of their models, making use of an idea proposed by Marx (1867) and Hicks
(1932).1 The argument is as follows: low unemployment and increasing bargaining power
of employees and their labour unions will speed up the increase in nominal and real wages,
which will finally generate a rising wage share and hence a falling profit share.2 This will
accelerate firms’ efforts to improve productivity growth in order to prevent the profit share
from falling. Dutt (2006) has recently argued that increasing pressure from lower
unemployment and rising real wages will accelerate the diffusion of innovations and will
thus increase productivity growth.
Taking into account both determinants the following equations for labour productivity
growth are obtained:
ŷ 5 h 1 ru 2 uh; h; r; u > 0;
ð10AÞ
736
E. Hein and A. Tarassow
which can be seen as including ‘learning by doing’ effects, which have been prominent in
older neoclassical models (Arrow, 1962).
3.3 Effects of a change in the profit share on the overall regime
Fig. 1. Growth equilibrium with endogenous productivity growth. (A) Capacity utilisation and
productivity growth. (B) Capital accumulation and productivity growth
1
See the Appendix for an analytical solution for the equilibrium and the effects of a change in distribution
on the equilibrium rates of capacity utilisation and productivity growth.
2
Also the profit rate (r 5 hu/v), which is not considered explicitly in this paper, is determined
endogenously.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
In order to calculate the total effect of a change in the profit share on demand and
productivity regime, we first have to determine the overall equilibrium with a given profit
share. Graphically, we obtain this equilibrium in Figure 1A, which contains the goods
market equilibrium rate of capacity utilisation from equation (8) and the productivity
equation (10A),1 and in Figure 1B, which shows the goods market equilibrium rate of
capital accumulation from equation (9) and the productivity equation (10B). With an
we obtain a dynamic equilibrium in which the rate of capacity
exogenous profit share (h),
utilisation (u**), the rate capital accumulation (g**) and the growth rate of labour
productivity (ŷ*) are determined endogenously.2 The ‘natural rate of growth’, or potential
growth, is hence endogenous in our model.
The existence and the stability of the overall equilibrium require that the slope of the
capacity utilisation equation (the capital accumulation equation) exceeds the slope of the
productivity equation in Figure 1A (in Figure 1B). Therefore, from equations (8) and
Distribution, aggregate demand and productivity growth
737
(10A) and Figure 1A we obtain the following condition for the existence and stability of an
overall equilibrium of capacity utilisation and productivity growth:
1
ð11Þ
½sW 1 ðsP 2 sW Þh 2 b 1 f 2 vr > 0:
v
Fig. 2. Increasing profit share and wage-led demand regime
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
From equations (9) and (10B) and Figure 1B the condition for existence and stability of an
overall equilibrium of capital accumulation and productivity growth is:
1
ð12Þ
ð1 2 veÞ ½sW 1 ðsP 2 sW Þh 1 f 2 b > 0:
v
Analysing the effects of a change in distribution on the overall equilibrium we have to
distinguish between wage-led and profit-led demand regimes. With a wage-led demand
regime the effects of a change in the profit share on aggregate demand, capital
accumulation, and productivity growth are in same direction: an increasing profit share
1 to h
2 ) has partially negative effects on the demand and on the productivity regime,
(from h
and these partial effects then reinforce each other. Figure 2 shows the total effect with
**
respect to capacity utilisation (reduction from u**
1 to u2 ) and productivity growth (decrease
*
*
from ŷ1 to ŷ2 ). Since we have not further differentiated the demand regimes regarding
different effects on capacity utilisation and capital accumulation, the overall effects on
capital accumulation and productivity growth are qualitatively basically the same and will
not be studied explicitly in what follows.
Under the conditions of a profit-led demand regime, a change in distribution has the
opposite effects on aggregate demand, capital accumulation and productivity growth. The
overall results of an increasing profit share will therefore depend on the relative strength of
each of these effects. If the expansive effect on the demand regime is rather weak, and the
contractive effect on the productivity regime is rather strong, we obtain an overall
contractive effect, as shown in Figure 3A: capacity utilisation, capital accumulation and
productivity growth are reduced. However, if the positive effect on the demand regime is
very pronounced and the negative effect on the productivity regime is rather weak, we
obtain an expansive overall case, as can be seen in Figure 3C: the rates of capacity
utilisation, capital accumulation and productivity growth increase in the face of a rising
profit share. With intermediate partial effects on demand and productivity regime, an
overall intermediate case is possible as well: an increasing profit share triggers higher rates
738
E. Hein and A. Tarassow
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Fig. 3. Increasing profit share and profit-led demand regime. (A) Contractive overall regime; (B)
intermediate overall regime; (C) expansive overall regime
of capacity utilisation and capital accumulation, but lower productivity growth rates, as is
displayed in Figure 3B. Table 2 summarises the potential effects of a changing profit share
on the demand, the productivity and the overall regime.
4. Review of the empirical literature
4.1 Distribution and aggregate demand
Recently, the effects of changes in functional income distribution on aggregated demand and
GDP, and hence the examination of the demand regime, have been subject of a range studies.1
1
For a more detailed survey of empirical results see Hein and Vogel (2008).
Distribution, aggregate demand and productivity growth
739
Table 2. Overall effects of a change in the profit share
Wage-led demand regime:
ð@u*=@hÞ < 0; ð@g*=@hÞ < 0
@u**=@h
@g**=@h
@ ŷ* @h
Overall regime
when profit share
is increasing
–
–
–
Contractive
Profit-led demand regime:
ð@u*=@hÞ > 0; ð@g*=@hÞ > 0
–
–
–
Contractive
1
1
–
Intermediate
1
1
1
Expansive
1
Short-run cyclical effects have been examined for instance by Barbosa-Filho and Taylor (2006) and
Stockhammer and Stehrer (2009) with mixed results.
2
Italy and Spain are also wage-led, according to Naastepad and Storm (2007), whereas Japan is found to
be profit-led.
3
For the Euro area, Stockhammer, Onaran and Ederer (2009) also find a wage-led demand regime for the
same period.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The focus of these studies has not been the short-run, cyclical effect of changes in
distribution on demand and GDP but rather medium to long run effects.1 Based on the
Bhaduri–Marglin (1990) model, these studies usually follow a method suggested by
Bowles and Boyer (1995) and estimate the effects of a change in functional income
distribution on aggregate demand applying a single equations estimation approach.
Changes in functional income distribution are assumed to be exogenous, as in the present
paper, and then the effects of these changes on consumption demand, investment demand
and net exports are estimated separately. Summing up these partial effects yields the total
effect of a change in income distribution on aggregate demand and GDP. Effects of
technical change on aggregate demand, however, are not considered in these studies.
For the period form the early 1960s to the mid 1980s, Bowles and Boyer (1995) obtain
a profit-led demand regime for France, Germany and Japan, whereas the UK and the USA
are wage-led in their view. Gordon (1995), however, finds the USA to be profit-led for
a similar period. For the period from 1960 to 2000, Naastepad and Storm (2007) derive
a wage-led demand regime for France, Germany, the Netherlands and the UK. The USA
are profit-led, according to their results.2 Naastepad (2006) and Ederer (2008) confirm the
result for the Netherlands in detailed country studies. For the period from 1960 to 2005,
Hein and Vogel (2008) find France, Germany, the UK, and the USA to be wage-led,
whereas Austria and the Netherlands are found to be profit-led. In more detailed country
studies, in particular with respect to the effects of redistribution on net exports, taking into
account some features of increasing globalisation, these results are confirmed by Stockhammer, Hein and Grafl (2009) for Germany and by Stockhammer and Ederer (2008) for
Austria.3 However, France is classified to be profit-led by Ederer and Stockhammer
(2007).
In order to take into account interdependencies between demand aggregates, which are
not included in the single equations estimation approach, Hein and Vogel (2009) started
with single equation estimations and then simulated the effects of increasing profit shares in
small multi-equations models for France and Germany in the period from 1960 to 2005. Their
results confirm a wage-led demand regime for France and also for Germany in the long run.
740
E. Hein and A. Tarassow
Although there do remain some differences regarding the classification of single
countries in the more recent studies, in particular regarding France, the Netherlands
and the USA, which may be due to differences in periods of investigation, data and
estimation methods, there appears, nonetheless, to arise a broad common result: Whereas
small open economies may be profit-led due to the dominance of the net export channel,
large and medium-sized open economies still tend to be wage-led. This seems to remain
true even in a period of increasing globalisation, increasing openness and increasingly
integrated markets, so that redistribution in favour of profits, although it is associated with
improved international competitiveness and rising net exports, will dampen the development of aggregate demand and GDP.
Estimations of the productivity regime can be broadly distinguished into those studies
estimating the effects of aggregate demand or capital accumulation on productivity growth
and those which estimate the effects of wage-push factors on productivity growth.
Recently, there have also been two studies that include both factors, besides other control
variables.
McCombie et al. (2002B) present an instructive survey on more than 80 studies, up to
2001, on the Verdoorn effect since the original study by Verdoorn (1949). They show that
the Verdoorn effect has been confirmed in the overwhelming majority of these studies with
different methods and data. This is true for cross-section estimations for countries or
regions (USA, UK and countries of the European Union, among others), or for industry
branches (USA, UK, France and Germany, among others), but also for time series
econometrics for single countries or regions (USA, UK and Germany, among others).
Therefore, McCombie (2002) summarises the results as follows:
In the three decades since the publication of the inaugural lecture there have been numerous
studies estimating the Verdoorn Law using a variety of different data sets. The picture that emerges
is, notwithstanding the instability of the law at the level of advanced countries and with some timeseries data sets, that the Verdoorn Law estimates are particularly robust with values of the Verdoorn
coefficient in the range of 0.3 to 0.6 and statistically significant. (McCombie, 2002, p. 106)
More recent studies, or studies not included in the McCombie et al. (2002B) overview,
confirm these results. Schnur (1990) estimates the Verdoorn equation for Germany from
1962 to 1988 and finds a significant elasticity of about 0.54 to 0.6 for the whole economy
and of around 0.4 for the manufacturing sector. Jasperneite and Allinger (1998) estimate
the typical Verdoorn equation for West Germany from 1988 to 1998 and find significant
coefficients ranging between 0.64 and 0.67. Walterskirchen (1999) presents some crosscountry analysis (1988–98) as well as country specific estimations (1970–98) for the USA,
Germany, the UK, France, the Netherlands and Austria, among others. He obtains
Verdoorn coefficients between 0.47 (USA) and 0.76 (Austria). Leon-Ledesma (2002) uses
some further control variables, like the investment share, the capital stock as a proxy of
innovations and the technology gap to the technological frontier, and gets a Verdoorn
coefficient about 0.64 and 0.67 for a cross-country regression containing 18 countries in
the period 1965–94. In a pooled regression for average values over the four business cycles
from 1960 to 1989 in 16 OECD countries (including Germany, the UK, the USA, the
Netherlands and Austria), Cornwall and Cornwall (2002) show that productivity growth
depends on the productivity gap to the USA (‘catching up factor’), on exports and
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
4.2 Aggregate demand, distribution and productivity growth
Distribution, aggregate demand and productivity growth
741
5. Estimation results for productivity growth
5.1 Data and estimation strategy
We applied single equation estimations in order to identify the effects of output growth and
distribution on productivity growth for six countries: Austria, France, Germany, the
Netherlands, the UK and the USA. We used annual data from the AMECO database
(European Commission, 2007), which cover a period from 1960 to 2007.2 Productivity
growth (ŷ) is the growth of real output per person employed (full-time equivalents) for the
economy as a whole. For the Verdoorn effect we used the growth rate of real GDP (Ŷ) as
a determinant of productivity growth. For the wage-push effect real wage growth (ŵ), as in
Naastepad (2006), was applied first. For the reasons given in the theoretical model, however,
we hold that the wage-push effect should be better indicated by the profit share. Therefore, we
estimated a second equation more consistent with our theoretical model with the profit
share (h) instead of real wage growth. Two control variables were included in order to avoid
the problem of unobserved variables. The share of manufacturing output in total GDP
1
See also Leon-Ledesma and Thirlwall (2002) who have shown for 15 OECD countries (including
France, Germany, the UK, the USA, the Netherlands and Austria) in the period 1960–95 that the natural
rate of growth, i.e. the sum of labour force growth and productivity growth, is positively affected by actual
GDP growth. The natural rate is thus endogenous with respect to the demand determined actual GDP
growth rate, with both productivity growth and labour supply growth being the endogeneity channels.
2
See the data description in the Appendix.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
investment demand, as well as the unemployment rate (‘demand factors’) and on percapita-income (indicator for structural change towards services). On average over the
countries, structural change towards services and the decreasing productivity gap to the
USA explain 50% of the productivity slowdown from the first sub-period (1960–73) to the
second (1974–89). The other 50% of the slowdown is explained by the demand factors.1
Finally, Uni (2007) finds rising effects of output growth on productivity growth for the US
(1978–87, 1988–2001) and the Japanese (1976–90, 1991–2003) manufacturing sectors
from the first to the second period estimated.
Wage-push factors for productivity growth are confirmed by Marquetti (2004). He finds
cointegration between real wages and labour productivity in the USA from 1869 to 1999.
Real wages are Granger-causal for labour productivity, but labour productivity is not
Granger-causal for real wages.
Naastepad (2006) in her country study for the Netherlands from 1960 to 2000 includes
a wage-push effect alongside the Verdoorn effect into the estimation of the productivity
equation. The elasticity of productivity growth with respect to real GDP growth is 0.63,
and with respect to real wage growth it is 0.52. Vergeer and Kleinknecht (2007) in paneldata regressions for 19 OECD countries from 1960 to 2004, including Germany, France,
the UK, the USA, the Netherlands and Austria, also estimate the effects of aggregate
demand and real wage growth on productivity growth. They include several other variables
into their estimations, such as the productivity gap to the technology leader (‘catching up’),
lagged labour productivity growth (‘path dependence’), the share of the service sector
(sectoral change), and country as well as time dummies. The estimated Verdoorn effect is
between 0.23 and 0.26, and the estimated wage push effect is between 0.24 and 0.34. In
the next section we first follow the approaches by Naastepad (2006) and Vergeer and
Kleinknecht (2007), but we then replace real wage growth as the cost-push variable by the
profit share in order to stick as close as possible to our theoretical model.
742
E. Hein and A. Tarassow
(sh_m) controls for the effects of structural change on economy-wide productivity growth.
Another impact controlled for is the potential for catching up with the technology leader.
Here we used the labour productivity difference between the USA and the respective
country under investigation (GAP). We therefore estimated the following general equations:
ð13Þ
ŷ 5 f Ŷ; ŵ; sh m; GAP ;
ð14Þ
ŷ 5 f Ŷ; h; sh m; GAP :
5.2 Estimation results
5.2.1. The real wage rate as cost-push variable. Starting with equation (13), an errorcorrection model could be estimated only for Germany:
d½logðyt Þ5a0 c 1 a1logðyt 2 1 Þ1a
t 2 1 Þ1a
1 1a5 logðGAPt 2 1 Þ
2 logðY
t 2 1 Þ1a3 logðw
4 sh mt 2
1 a6 +d log yt 2 j 1a7 +d log Yt 2 i 1 a8 + d log wt 2 i 1a9 +d½sh mt 2 i 1 a10 +d½logðGAPt 2 i Þ 1 et ;
i 5 1; 2; 3; j 5 1; 2
ð13AÞ
For the other countries we estimated dynamic difference equations of the following form:
d log yt 5 b0 c 1 b1 +d log yt 2 j 1 b2 +d log Yt 2 i 1 b3 +d½logðwt 2 i Þ
ð13BÞ
1 b4 +d sh mt 2 i 1 b5 +d½logðGAPt 2 i Þ 1 et ;
i 5 0; 1; 2; 3; j 5 1; 2
The coefficients (a2) and (b2) are the Verdoorn coefficients, the coefficients (a3) and (b3)
are the wage-push coefficients. The long-run coefficients of the ECM were calculated by
dividing the level variable of the exogenous variable by the absolute value of the errorcorrection term (e.g. a2/|a1|). The long-run coefficients for the difference equations were
obtained by summing up the coefficients of the exogenous variables and dividing them by
one minus the coefficient of the lagged endogenous variable [e.g. b3/(1–b1)]. Table 3 shows
the estimation results.
1
2
3
ADF and ADF-GLS tests are available from the authors on request.
See Elliot et al. (1992) for the construction of this test.
See also the critique in Vergeer and Kleinknecht (2007).
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The estimation strategy was as follows. All level variables were transformed into logs. First
we tried to estimate an error-correction-model (ECM) following the method suggested by
Pesaran et al. (2001). They simulated critical value bounds for an F test in order to determine
long-run relationships, independently of the degree of integration of the respective
variables.1 If no cointegration was found, dynamic difference models with lags up to
three periods were estimated. Applying the ADF test, some variables seemed to be I(0),
but the ADF-GLS test2 could not reject the null hypothesis of the existence of a unit root, so
that we included the variables in first differences. Insignificant variables were excluded
and the equations were re-estimated. All necessary tests were applied. The lags in the
difference equations were chosen in order to cope with cyclical variations in the variables and
to capture long-run effects. Thus, some effects in our study—especially the Verdoorn
effect—seem to be relatively small compared with other studies reviewed in the previous
section, because in most of those studies only the contemporaneous effect is considered.3
Table 3. Determinants of productivity growth in Germany, France, the Netherlands, Austria, the UK and the USA, 1960–2007
Endogenous: d[log(y)]
Constant
p value
log(Yt–1)
p value
log(yt–1)
p value
log(wt–1)
p value
sh_mt–1
p value
log(GAPt–1)
p value
d[log(yt–1)]
p value
d[log(Yt)]
p value
d[log(Yt–1)]
p value
d[log(Yt–2)]
p value
d[log(Yt–3)]
p value
d[log(Yt–4)]
p value
d[log(wt)]
p value
d[log(wt–1)]
p value
d[log(wt–3)]
–0.17
0.30
0.12
0.02
–0.28
0.00
0.09
0.04
–0.10
0.51
0.04
0.00
–0.35
0.00
–0.29
0.02
France
0.74
0.00
0.60
0.00
–0.69
0.00
0.10
0.04
0.13
0.00
Netherlands
0.58
0.00
0.63
0.00
–0.65
0.00
Austria
UK
USA
0.00
0.95
0.02
0.00
0.01
0.00
0.51
0.00
0.20
0.02
0.56
0.00
–0.47
0.00
0.76
0.00
0.76
0.00
–0.81
0.00
0.13
0.00
0.21
0.00
0.14
0.00
0.16
0.00
–0.58
0.00
0.12
0.01
0.18
0.00
0.25
0.00
0.29
0.00
0.08
743
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Germany
Distribution, aggregate demand and productivity growth
1960–2007
744
Table 3. Continued
Endogenous: d[log(y)]
p value
d(sh_mt)
p value
d(sh_mt–1)
p value
d(sh_mt–2)
p value
d(sh_mt–3)
p value
d[log(GAPt)]
p value
d[log(GAPt–2)]
p value
(dy/y)/(dY/Y)
(dy/y)/(dw/w)
Adjusted R2
D-W statistics
Reset test, p value
White’s test, p value
Breusch–Pagan, p
value
Normal distribution,
p value
LM test(3), p value
CUSUM, p value
Wald test, F stat
Dummies and
determinants
Germany
France
Netherlands
Austria
UK
USA
0.05
–0.53
0.00
0.46
0.02
0.76
0.02
–1.02
0.00
0.39
0.00
0.41
0.00
–0.93
0.00
–0.02
0.09
0.43
0.32
0.71
2.10
0.37
0.92
0.97
0.54
0.31
0.98
1.97
0.70
0.19
0.77
0.45
0.33
0.97
2.22
0.34
0.25
0.43
0.03
0.05
0.33
0.67
0.96
1.80
0.79
0.40
0.63
0.60
0.57
0.66
0.76
0.57
0.17
0.21
0.18
0.24
0.65
0.69
0.76
0.66
0.53
0.47
0.68
0.26
0.24
19.22***
Dummy 1976
Dummy 1979 and
1981
Dummy 1966 and
1982
0.23
0.25
0.90
1.88
0.18
0.35
0.64
0.11
0.36
0.89
1.68
0.98
0.29
0.63
Dummy 1974 and
1981, time trend
Notes: y, labour productivity; Y, GDP; w, real wage; sh_m, share of manufacturing sector; GAP, labour productivity gap with the USA.
Dummy 1963, 1976,
1978 and 1984
E. Hein and A. Tarassow
1960–2007
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Distribution, aggregate demand and productivity growth
745
5.2.2. The profit share indicating cost-push effects. Having so far confirmed the results
from the literature (i.e. Naastepad, 2006; Vergeer and Kleinknecht, 2007), in the next step
we estimated the, in our view, more appropriate productivity equation (14). Inspecting the
data, however, reveals that there is a structural break in the relationship between the profit
share and productivity growth in the early/mid 1980s, except for the UK and the USA
(Figure 4).1
Considering this aspect, we decided to estimate sub-periods to account for this
structural break for Austria, France, Germany and the Netherlands. For the USA
and UK we estimated the whole period. For both countries we could not find any
cointegration relationship. So we had to estimate difference models for all
countries:
d log yt 5 b0 c 1b1 +d log
yt 2 j 1 b2 +d log Yt 2 i 1 b3 +d½ht 2 i ð14AÞ
1 b4 +d sh mt 2 i 1 b5 +d½logðGAPt 2 i Þ 1 et ;
i 5 0; 1; 2; 3; j 5 1; 2
Table 4 for the UK and the USA shows that the R2 are relatively high and the models are
robust. For the UK the null hypothesis of the Breusch–Pagan test can only be rejected at
10% significance, but the White test confirms homoskedasticity of the residuals. The
Verdoorn effects are higher than for the first estimation, due to the fact that for both models
only current GDP growth is significant. However, we can again confirm the generally
1
Applying the QLR test as well as the Chow test in order to determine structural breaks does not obtain
any appropriate results. This could be due to the fact that other structural breaks, as in 1979 and 1982, overcompensate the structural break in the mid 1980s. But the sub period specific trends in Figure 4 show a clear
switch.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The long-run coefficients were highly significant and show the expected sign. The
explanatory content (R2) is relatively high. When necessary, the models were corrected
for outliers in order to prevent heteroskedasticity. The test results verify robust models.
For Germany the Wald test can reject the null that the level coefficients are different
from zero at 1% significance. Hence, there exists a long-run equilibrium between the
variables.
The estimation results for the Verdoorn effect are generally smaller compared with other
studies. This difference may be due to the introduction of lagged variables in our approach,
i.e. other studies might have captured short-run cyclical Okun effects rather than the
Verdoorn effect, as Vergeer and Kleinknecht (2007) have recently argued. The strongest
influences of output growth on productivity growth were found for France and the lowest
for the USA. A 1% increase of output raises labour productivity by 0.54% in France. The
corresponding results for Germany (0.43%) and the Netherlands (0.45%) are of similar
magnitude; for Austria and the UK we have found smaller values (0.33% and 0.23%,
respectively) and for the USA the long-run output elasticity of labour productivity is only
0.11%.
The effect of real wages on labour productivity is significantly positive. The elasticities
are in the range of other studies, but differ between countries. The highest values were
found for Austria. A 1% increase in real wages raises labour productivity by approximately
0.67%. The values for the Netherlands and the USA are of similar size (0.33% and 0.36%,
respectively). The elasticities for Germany (0.32%), France (0.31%) and the UK (0.25%)
are lower.
746
E. Hein and A. Tarassow
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Fig. 4. Profit share and labour productivity growth, 1960–2007 for (A) Germany, (B) France, (C)
the Netherlands, (D) Austria, (E) the UK and (F) the USA
known results. The contemporaneous relations for the UK report that an increase in
output by 1% raises labour productivity by 0.61%. In the USA this effect is about 0.39%.
The distributional effects are in accordance with our theoretical hypothesis: an increase in
Distribution, aggregate demand and productivity growth
747
Table 4. Determinants of productivity growth in the UK and the USA, 1960–2007
Endogenous: d[log(y)]
USA
0.01
0.00
0.61
0.00
–0.46
0.00
0.00
0.12
0.39
0.00
–0.33
0.02
–1.53
0.00
0.21
0.11
–0.08
0.01
0.61
–0.46
0.69
1.65
0.39
0.29
0.09
0.48
0.66
0.27
Dummy 1988
0.39
–0.33
0.73
1.67
0.40
0.93
0.92
0.63
0.57
0.52
Dummy 1964, 1979, 1987 and 1992
Notes: y, labour productivity; Y, GDP; h, adjusted profit share; sh_m, share of manufacturing sector; GAP,
labour productivity gap to USA.
the profit share reduces labour productivity growth. A one percentage point increase in the
profit share decreases labour productivity by 0.46% in the UK and by about 0.33% in the
USA.
The models for the other countries are also robust, as can be seen in Table 5. For France
and the Netherlands the Durbin–Watson statistics are very low in the first sub-period, but
the LM test cannot reject the null hypothesis that there is no autocorrelation for the first as
well as for the third lag.1 The same applies for Germany, the Netherlands and Austria in
the second sub-period. In both periods and for all countries the R2 values are relatively
high. All tests verify the efficiency and stability of the equations. Both periods show
a significant influence of the Verdoorn effect. The wage-push effect, measured by the profit
share, does only seem to hold for the second period, with the exception of France where we
have the expected sign but no statistical significance.
Generally we found that the Verdoorn effect is stronger in the first sub-period than in the
second, except for Austria. In Germany, this decline is most pronounced, from 0.86 in the
first period to 0.27 in the second.
1
Since we included lags of the endogenous variable, the Durbin–Watson statistics is biased (see
Kirchgässner and Wolters, 2006, p. 17).
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Constant
p value
d[log(Yt)]
p value
d(ht–2)
p value
d(ht–3)
p value
d(sh_mt–1)
p value
d(sh_mt–2)
p value
d[log(GAPt)]
p value
(dy/y)/(dY/Y)
(dy/y)/(dh)
Adjusted R2
D-W statistics
Reset test, p value
White’s test, p value
Breusch–Pagan, p value
Normal distribution, p value
LM test(3), p value
CUSUM, p value
Dummies and determinants
UK
748
Table 5. Determinants of productivity growth in Germany, France, the Netherlands and Austria, 1960–2007 (sub-periods)
Endogenous: d[log(y)]
1960–
1984
0.59
0.00
–0.35
0.04
0.72
0.00
0.80
0.00
–0.71
0.00
Netherlands
Austria
1985–
2007
1960–
1982
1983–
2007
1960–
1983
1984–
2007
1960–
1983
1985–
2007
0.01
0.03
0.13
0.08
0.01
0.00
0.70
0.00
0.03
0.00
0.36
0.00
0.01
0.00
0.66
0.00
0.01
0.04
0.27
0.01
0.05
0.00
0.32
0.00
0.01
0.00
0.48
0.00
–0.18
0.02
0.32
0.00
0.52
0.00
0.29
0.01
–0.42
0.01
0.15
0.01
–0.07
0.69
–0.1
0.59
0.67
0.00
–0.46
0.00
–0.33
0.00
0.37
0.05
–0.98
0.00
–0.34
0.10
–0.07
0.00
0.86
0.27
0.03
0.01
0.70
0.36
–0.05
0.03
0.66
0.27
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Const
p value
d[log(Yt)]
p value
d[log(Yt–1)]
p value
d[log(yt–1)]
p value
d(ht)
p value
d(ht–1)
p value
d(ht–2)
p value
d(sh_mt)
p value
d(sh_mt–1)
p value
d(sh_mt–2)
p value
d[log(GAPt–1)]
p value
d[log(GAPt–2)]
p value
(dy/y)/(dY/Y)
France
0.32
0.44
E. Hein and A. Tarassow
Germany
Table 5. Continued
Endogenous: d[log(y)]
Germany
Netherlands
Austria
1960–
1984
1985–
2007
1960–
1982
1983–
2007
1960–
1983
1984–
2007
1960–
1983
1985–
2007
0.32
0.96
2.24
0.98
0.23
0.42
–0.87
0.85
2.46
0.85
0.47
0.72
0.15
0.96
1.51
0.78
0.85
0.34
–
0.56
1.82
0.22
0.43
0.18
0.29
0.90
1.60
0.81
0.48
0.80
–0.33
0.60
2.48
0.70
0.53
0.42
0.67
0.94
1.97
0.98
0.31
0.47
–0.68
0.91
2.45
0.75
0.25
0.13
0.85
0.98
0.19
0.61
0.83
0.39
0.79
0.52
0.87
0.64
0.53
0.12
Dummies
2005 and
2006
0.48
0.59
Dummy
1968
0.96
0.99
Dummy
2001, time
trend
0.65
0.20
Dummies
1979 and
1980
0.26
0.26
Dummies
1984 and
2004
0.40
0.83
Dummy 1965,
time trend
0.58
0.96
Dummy
1996
Notes: y, labour productivity; Y, GDP; h, adjusted profit share; sh_m, share of manufacturing sector; GAP, labour productivity gap to USA.
Distribution, aggregate demand and productivity growth
(dy/y)/(dh)
Adjusted R2
D-W statistics
Reset test, p value
White’s test, p value
Breusch–Pagan, p
value
Normal distribution,
p value
LM test(3), p value
CUSUM, p value
Dummies and
determinants
France
749
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
750
E. Hein and A. Tarassow
For distribution we found a significantly positive effect of the profit share on productivity
growth in all four countries in the first period, which is hardly to square with our theoretical
arguments. In the second sub-period, from the early/mid 1980s until 2007, however, the
expected adverse relationship between the profit share and productivity growth was
obtained, except for the statistical significance problem with France. In Germany, an
increase in the profit share by one percentage point has decreased productivity by 0.87%.
In Austria the effect was 0.68% and in the Netherlands it was 0.33%. The change in the
relationship between distribution and productivity growth in the continental European
countries remains to be assessed in further research. Here we can only speculate that the
relationship between the profit share and productivity growth may be non-linear as, for
example, Lima (2004) has suggested.1
Starting from a simple theoretical model for an open economy with productivity growth we
have examined the effects of a change in functional income distribution on the demand, the
productivity and the overall regime of the economy. Generally, we obtain that with
endogenous productivity growth potential GDP growth becomes path dependent with
respect to changes in distribution. With wage-led aggregate demand the negative effects of
an increasing profit share on the demand and the productivity regime reinforce each other
and an overall contractive regime emerges. If aggregate demand is profit-led, however,
different overall regimes may arise in the face of an increasing profit share, depending on
the relative strength of the effects of redistribution on the demand and on the productivity
regime: contractive or expansive effects throughout on capacity utilisation, capital
accumulation and productivity growth, or an intermediate regime with positive effects
on economic activity and capital accumulation, but negative effects on productivity
growth.
Recent empirical studies imply that the medium- to long-run demand regime in large
and medium-sized open economies, as in Germany, France, the UK and the USA, tends to
be wage-led, whereas for small open economies, as the Netherlands and Austria, some
studies have obtained profit-led results. Our estimations of the productivity regime for
these six countries in the period from 1960 to 2007 have confirmed the prevalence of
a Verdoorn effect, i.e. a positive impact of GDP growth on productivity growth. In
countries with wage-led demand regimes, therefore, redistribution at the expense of labour
not only weakens aggregate demand and GDP growth but, through the Verdoorn effect,
productivity growth is also negatively affected. We have also introduced wage-push
variables into the estimations of the productivity equations. A positive effect of real wage
growth on productivity growth has been confirmed for all six countries. However, since
only real wage growth exceeding productivity growth eats into unit profits and the profit
share imposing cost-cutting pressure on firms, we have replaced real wage growth by the
profit share in the estimations of the productivity equations, consistent with our theoretical
model. A negative effect of the profit share on productivity growth was found for the UK
and the USA for the whole period, and for Austria, Germany and the Netherlands for the
period from the early/mid 1980s to 2007. For France no statistically significant effect could
1
In Lima’s (2004) model, the profit share has two-fold effects on productivity growth: it affects the
incentive to innovate, as in our model, but it also affects the funds to innovate. However, he has no effect of
demand on productivity growth in his model.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
6. Conclusions
Distribution, aggregate demand and productivity growth
751
Bibliography
Arrow, K. 1962. The economic implications of learning by doing, Review of Economic Studies, vol.
29, 155–73
Barbosa-Filho, N. H. and Taylor, L. 2006. Distributive and demand cycles in the US
economy—a structuralist Goodwin model, Metroeconomica, vol. 57, 389–411
Bhaduri, A. 2006. Endogenous economic growth: a new approach, Cambridge Journal of
Economics, vol. 30, 69–83
Bhaduri, A. and Marglin, S. 1990. Unemployment and the real wage: the economic basis for
contesting political ideologies, Cambridge Journal of Economics, vol. 14, 375–93
Blecker, R. A. 1989. International competition, income distribution and economic growth,
Cambridge Journal of Economics, vol. 13, 395–412
Bowles, S. and Boyer, R. 1995. Wages, aggregate demand, and employment in an open economy:
an empirical investigation, in Epstein, G. A. and Gintis, H. M. (eds), Macroeconomic Policy after
the Conservative Era, Cambridge, Cambridge University Press
Cassetti, M. 2003. Bargaining power, effective demand and technical progress: a Kaleckian
model of growth, Cambridge Journal of Economics, vol. 27, 449–64
Cornwall, J. and Cornwall, W. 2002. A demand and supply analysis of productivity growth,
Structural Change and Economic Dynamics, vol. 13, 203–29
Dutt, A. K. 2003. New growth theory, effective demand, and post-Keynesian dynamics, in
Salvadori, N. (eds), Old and New Growth Theories: An Assessment, Cheltenham, Edward Elgar
Dutt, A. K. 2006. Aggregate demand, aggregate supply and economic growth, International
Review of Applied Economics, vol. 20, 319–36
Ederer, S. 2008. ‘Competition-oriented Wage Policies and its Effects on Effective Demand in the
Netherlands’, WIFO Working Papers 312/2008, Vienna, WIFO
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
be found. Through the wage-push channel a falling wage share also has a directly negative
impact on productivity growth.
Summing up, in those wage-led economies in which we also found a statistically
significant negative effect of the profit share on productivity growth, as in Germany, the
UK and the USA, the dampening effects of a rising profit share on the demand and the
productivity regime reinforce each other and an overall contractive effect of a rising profit
share emerges. In France with a wage-led demand regime but no significant direct effect of
the profit share on productivity growth, we also obtain an overall contractive effect of a rise
in the profit share, because the negative aggregate demand effect also spills over to
productivity growth via the Verdoorn channel.
In countries with a profit-led demand regime, however, as probably in the Netherlands
or in Austria, the expansive effects of an increasing profit share on aggregate demand go
along with a partially depressive effect on productivity growth, which, however, may be
compensated for by the expansive effect via GDP growth and the Verdoorn effect.
Therefore, the character of the overall regime in these countries (contractive, intermediate
or expansive) remains to be determined in more detailed empirical analysis, which will have
to focus on the interaction of demand and productivity regime.
In those economies with a wage-led demand regime, however, a clear-cut result for
economic policies is obtained from our analysis, at least for the period since the mid 1980s.
Redistribution at the expense of labour is not only harmful for aggregate demand and
economic activity in the short run, it has depressing effects on capital accumulation and
productivity growth, and hence on potential growth and the ‘natural rate of growth’, in the
long run. If potential feedbacks on distribution, not discussed in the present paper, were
included, the danger of cumulative downwards processes in these economies are quite
obvious, with falling wage shares, falling GDP, capital stock and productivity growth rates
reinforcing each other.
752
E. Hein and A. Tarassow
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
Ederer, S. and Stockhammer, E. 2007. Wages and aggregate demand: an empirical investigation
for France, in Hein, E. and Truger, A. (eds), Money, Distribution and Economic
Policy—Alternatives to Orthodox Macroeconomics, Cheltenham, Edward Elgar
Elliot, G., Rothenberg, T. J. and Stock, J. H. 1992. ‘Efficient Tests for an Autoregressive Unit
Root’, NBER Technical Working Paper Series No. 130, Cambridge, NBER
European Commission. 2007. Annual macro-economic database (AMECO), November,
available at http://ec.europa.eu/economy_finance/db_indicators/db_indicators8646_en.htm
[date last accessed 12 April 2008]
Foley, D. K. and Michl, T. R. 1999. Growth and Distribution, Cambridge, MA, Harvard
University Press
Gordon, D. 1995. Growth, distribution, and the rules of the game: social structuralist
macro foundations for a democratic economic policy, in Epstein, G. A. and Gintis, H. M.
(eds), Macroeconomic Policy after the Conservative Era, Cambridge, Cambridge University
Press
Hein, E. 2005. Löhne, Verteilung und Wachstum: Ansätze in der Tradition Michal Kaleckis, in
Hein, E., Heise, A., and Truger, A. (eds), Löhne, Beschäftigung, Verteilung und Wachstum.
Makroökonomische Analysen, Marburg, Metropolis
Hein, E. 2008. Money, Distribution Conflict and Capital Accumulation. Contributions to
‘Monetary Analysis’, Basingstoke, Palgrave Macmillan
Hein, E. and Vogel, L. 2008. Distribution and growth reconsidered—empirical results for six
OECD countries, Cambridge Journal of Economics, vol. 32, 479–511
Hein, E. and Vogel, L. 2009. Distribution and growth in France and Germany—single equation
estimations and model simulations based on the Bhaduri/Marglin-model, Review of Political
Economy, vol. 21, 245–72
Hein, E., Schulten, T. and Truger, A. 2006. Deflation risks in Germany and the EMU: the role of
wages and wage bargaining, in Hein, E., Heise, A. and Truger, A. (eds), Wages, Employment,
Distribution and Growth. International Perspectives, Basingstoke, Palgrave Macmillan
Hicks, J. 1932. The Theory of Wages, London, Macmillan
Jasperneite, C. and Allinger, H. J. 1998. ‘Langfristige Perspektiven am westdeutschen
Arbeitsmarkt: Was sagen die Gesetze von Okun und Verdoorn?’, Discussion Paper of the
University of Passau, 12/1998, 2nd edn
Kaldor, N. 1957. A model of economic growth, The Economic Journal, vol. 67, 591–624
Kaldor, N. 1961. Capital accumulation and economic growth, in Lutz, F. A. and Hague, D. C.
(eds), The Theory of Capital, London, Macmillan
Kaldor, N. 1966. Causes of the Slow Rate of Economic Growth in the United Kingdom, Cambridge,
Cambridge University Press
Kalecki, M. 1954. Theory of Economic Dynamics, London, George Allen and Unwin
Kirchgässner, G. and Wolters, J. 2006. Einführung in die moderne Zeitreihenanalyse, Munich,
Franz Vahlen
Kurz, H. D. 1991. Technical change, growth and distribution, in Nell, E. and Semmler, W. (eds),
Nicholas Kaldor and Mainstream Economics: Confrontation or Convergence?, London, Macmillan
Leon-Ledesma, M. A. 2002. Accumulation, innovation and catching-up: an extended
cumulative growth model, Cambridge Journal of Economics, vol. 25, 201–16
Leon-Ledesma, M. A. and Thirlwall, A. P. 2002. The endogeneity of the natural rate of growth,
Cambridge Journal of Economics, vol. 26, 441–59
Lima, G. T. 2000. Market concentration and technological innovation in a dynamic model of
growth and distribution, Banca Nazionale del Lavoro Quarterly Review, vol. 53, 447–75
Lima, G. T. 2004. Endogenous technological innovation, capital accumulation and
distributional dynamics, Metroeconomica, vol. 55, 386–408
Marquetti, A. 2004. Do rising real wages increase the rate of labor-saving technical change?
Some econometric evidence, Metroeconomica, vol. 55, 432–41
Marx, K. 1867. Das Kapital. Kritik der politischen Ökonomie, Erster Band: Der Produktionsprozeß
des Kapitals, 4th, editon 1890, edited by Engels, F., Marx-Engels-Werke, Volume 23, Berlin
1962, Dietz Verlag
McCombie, J. 2002. Increasing returns and the Verdoorn Law from a Kaldorian perspective, in
McCombie, J., Pugno, M. and Soro, B. (eds), Productivity Growth and Economic Performance.
Essays on Verdoorn’s Law, Basingstoke, Palgrave Macmillan
Distribution, aggregate demand and productivity growth
753
Appendix
A1 Solutions for overall equilibrium and for overall effects of a change in the
profit share on capacity utilisation and productivity growth
From equations (10A) and (8) the overall equilibrium rates of capacity utilisation and
productivity growth are obtained:
u** 5
a 1 ðt 2 uvÞh 1 cer ðhÞ 1 vh
:
½sW 1 ðsP 2 sW Þh1v 2 b 1 f 2 vr
ðA1Þ
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
McCombie, J., Pugno, M. and Soro, B. (eds), 2002A. Productivity Growth and Economic
Performance. Essays on Verdoorn’s Law, Basingstoke, Palgrave Macmillan
McCombie, J., Pugno, M. and Soro, B. 2002B. Introduction, in McCombie, J., Pugno, M., and
Soro, B. (eds), Productivity Growth and Economic Performance. Essays on Verdoorn’s Law,
Basingstoke, Palgrave Macmillan
Naastepad, C. W. M. 2006. Technology, demand and distribution: a cumulative growth model
with an application to the Dutch productivity growth slowdown, Cambridge Journal of
Economics, vol. 30, 403–34
Naastepad, C. W. M. and Storm, S. 2007. OECD demand regimes (1960–2000), Journal of Post
Keynesian Economics, vol. 29, 211–46
Pesaran, H. M., Shin, Y. and Smith, R. J. 2001. Bounds testing approaches to the analysis of level
relationships, Journal of Applied Econometrics, vol. 16, 289–326
Rowthorn, R. 1981. Demand, real wages and economic growth, Thames Papers in Political
Economy, Autumn
Schnur, P. 1990. Investitionstätigkeit und Produktivitätsentwicklung: empirische Analyse auf der
Basis eines erweiterten Verdoorn-Ansatzes, Mitteilungen aus der Arbeitsmarkt- und
Berufsforschung, vol. 23, 106–10
Setterfield, M. and Cornwall, J. 2002. A neo-Kaldorian perspective on the rise and the decline of
the Golden Age, in Setterfield, M. (ed.), The Economics of Demand-led Growth. Challenging the
Supply-side Vision of the Long Run, Cheltenham, Edward Elgar
Stockhammer, E. and Ederer, S. 2008. Demand effects of the falling wage share in Austria,
Empirica, vol. 35, 481–502
Stockhammer, E., Hein, E. and Grafl, L. 2009. Globalization and the effects of changes in
functional income distribution on aggregate demand in Germany, International Review of
Applied Economics, forthcoming
Stockhammer, E., Onaran, Ö and Ederer, S. 2009. Functional income distribution and
aggregate demand in the Euro area, Cambridge Journal of Economics, vol. 33, 139–59
Stockhammer, E. and Stehrer, R. 2009. Goodwin or Kalecki in demand? Functional income
distribution and aggregate demand in the short run, Political Economy Research Institute Working
Paper, Number 203, University of Massachusetts, Amherst
Taylor, L. 1991. Income Distribution, Inflation, and Growth. Lectures on Structuralist Macroeconomic
Theory, Cambridge, MA, London, MIT Press
Uni, H. 2007. Growth regimes in Japan and the United States in the 1990s, Revue de la
Regulation, no. 1, 1–20
Verdoorn, P. J. 1949. Factors that determine the growth of labour productivity, in McCombie, J.,
Pugno, M., and Soro, B. (eds), Productivity Growth and Economic Performance. Essays on
Verdoorn’s Law, 2002, Basingstoke, Palgrave Macmillan
Vergeer, R. and Kleinknecht, A. 2007. ‘Jobs Versus Productivity? The Causal Link from Wages to
Labour Productivity Growth’, TU Delft Innovation Systems Discussion Papers, IS 2007–01
Walterskirchen, E. 1999. ‘The relationship between growth, employment and unemployment in
the EU’, paper presented at a Workshop in Barcelona, 16–18 September 1999, European
Economists for an Alternative Economic Policy (TSER Network)
You, J.-I. 1994. Macroeconomic structure, endogenous technical change and growth, Cambridge
Journal of Economics, vol. 18, 213–33
754
E. Hein and A. Tarassow
ðh 2 uhÞ ½sW 1 ðsP 2 sW Þh1v 2 b 1 f 1 r½a 1 th 1 cer ðhÞ
ŷ* 5
;
½sW 1 ðsP 2 sW Þh1v 2 b 1 f 2 vr
ðA2Þ
The effect of a change in the profit share on the overall equilibrium rate capacity utilisation
is given by:
r
t 2 ðsP 2 sW Þuv 1 c@e
@u**
@h 2 uv
5
:
@h
½sW 1 ðsP 2 sW Þh1v 2 b 1 f 2 rv
ðA10 Þ
As can easily be seen, the total effect of a change in the profit share is composed of two subeffects.1 The effect via goods market activity { r½t2ðsP 2sW Þðu=vÞ1cð@er =@hÞ} may be
positive or negative depending on the nature of the demand regime. If demand is profit-led,
this effect will be positive, if it is wage-led, this effect will be negative. The second effect
(2uf½sW 1ðsP 2sW Þhð1=vÞ2b1fg) is negative in any case, because the term in brackets
has to be positive from the goods market stability condition. This term captures the directly
negative effect of an increase in the profit share on productivity growth via the cost-push
channel. Therefore, in a wage-led demand regime, the overall effect of an increasing profit
share on productivity growth will be negative, whereas in a profit-led demand regime the
overall effect of a rising profit share on productivity growth may be either positive or
negative.
A2 Data definitions and data source
All variables are obtained from the AMECO database (European Commission, 2007):
y: gross domestic product at 2000 market prices per person employed (full-time
equivalents)
Y: gross domestic product at 2000 market prices
w: real compensation per employee, deflator private consumption: total economy
h: 1 minus adjusted wage share (total economy), as percentage of GDP at factor cost
(wage share: compensation per employee as percentage of GDP at factor cost per person
employed)
sh_m: share of gross value added in manufacturing industry in gross domestic product at
2000 market prices
GAP: difference of GDP at 2000 market prices per person employed (full-time
equivalents) with respect to the USA (for all countries in levels and euro)
1
Again, the denominator is positive from the existence and stability condition of overall equilibrium.
Downloaded from http://cje.oxfordjournals.org by on July 12, 2010
The denominator has to be positive from the existence and stability condition of the overall
equilibrium [equation (11)]. There are positive effects of an increasing profit share via
investment (t) and net exports [cð@er =@hÞ)], a negative effect via consumption
[2ðsP 2sW Þðu=vÞ], and now also negative effects via productivity growth (2uv). The
overall effect may hence be positive (profit-led) or negative (wage-led), depending on the
strength of the individual effects.
For the effect of a change in the profit share on the equilibrium rate of productivity
growth we obtain:
1
r
@ ŷ* r t 2 ðsP 2 sW Þuv 1 c@e
@h 2 u ½sW 1 ðsP 2 sW Þhv 2 b 1 f
5
:
ðA20 Þ
@h
½sW 1 ðsP 2 sW Þh1v 2 b 1 f 2 rv